基于SVM的模糊推理在图像降噪中的建模与仿真

针对图像上的脉冲曝声，用基于支持向量机的模糊推理方法建立噪声检测模型。该建模方法应用支持向量机的学习机制从训练样本中提取支持向量，由支持向量确定模糊基函数，产生相应的模糊规则，建立起模糊推理模型。并依据此设计了一套噪声检测系统。该系统由基于支持向量机的模糊推理子系统和决策子系统组成。其中，推理子系统分别在纵向和横向上检测噪声信息：决策子系统综合纵向横向的信息，做出决策。仿真实验结果表明，所提出的方法可有效地检测并去除噪声，同时保留了图像的细节信息。

Modeling and Simulation of SVM-based Fuzzy Inference System for Image Denoise

The purpose of this paper is to build up a detection model by using SVM-based fuzzy inference system for the impulse noise in an image. The support vectors which are used for confirming the fuzzy basis function and creating the corresponding fuzzy rules were extracted from the training samples by the learning mechanism of SVM, and the fuzzy reasoning model was build up as a result. Accordingly, a noise detection system was constructed. This system is composed of two SVM-based fuzzy inference subsystems and a decision subsystem. Therein, the inference subsystems detect noise information on horizontal and vertical orientation, respectively; and the decision subsystem made decision by synthesizing horizontal and vertical information. The experiment results show that the proposed method can detect correctly and remove noise while preserving the detail information.